The first system described for the successful in vitro exponential amplification of target nucleic acids is the Polymerase Chain Reaction (PCR) (Saiki et al., 1985 Science 230; 1350-1354). PCR has been widely used for allele determination, forensic identification, gene analysis, diagnostics, cloning, direct sequencing and other applications. Subsequently, Reverse Transcriptase (RT) was used to transform an RNA molecule into a DNA copy allowing the use of RNA molecules as substrates for PCR amplification by DNA polymerase. In addition, conditions have been described that allow certain DNA polymerases to perform reverse transcription by themselves (Myers, T. W. and Gelfand, D. H. [1991] Biochem. 30; 7661-7666), contents incorporated herein by reference. Finally, Rose et al. (U.S. Pat. No. 5,508,178, also incorporated herein by reference) have described the use of inverted repeat sequences as choices for PCR primer sequences, allowing the use of a single primer to initiate polymerization from each end of a target nucleic acid to create a PCR amplicon that in single-stranded form can be drawn as a “pan-handles” with self complementary sequences at each end. In order to utilize targets that lack inverted repeats, this group has also described various methods to introduce sequences into a PCR amplicon, such that the final product would have self-complementary sequences at each end (U.S. Pat. Nos. 5,439,793, 5,595,891, and 5,612,199, each of which is incorporated herein by reference).
Both the original PCR amplification system and various improved PCR systems suffer from the limitation of a necessity for expensive dedicated thermocyclers to provide the multiple temperature conditions that are intrinsic to the PCR method. This necessity is derived from the problem that the extension of a primer creates a product that has a stronger association with a template than the primer used to create it. As such, in a system like PCR, temperatures that allow binding of a primer are too low to allow separation of the extended product from its template and temperatures that are elevated enough to allow the separation of the extended product are too high to allow another priming event. The second priming event cannot take place until after the first extended strand is separated from its template. As such, in PCR amplification, primer binding to template and the sequential release of the extended primers from the template have to be carried out at separate distinct temperatures and require a thermocycler to provide repeated sequences of distinct thermal steps. The existence of discrete cycles with different conditions also necessitates an optimization of temperature for each individual temperature step as well as an appropriate timing for each step. Similar problems also apply when ligation is used in the LCR reaction (Backman, K. et al. European Patent Application Publication No. 0 320 308, Landegren, U., et al., 1988 Science 241; 1077, Wu, D. and Wallace, R. B. 1989 Genomics 4; 560, Barany, F. 1991 Proc. Nat. Acad, Sci. USA 88; 189) where the temperature required for binding individual probes is less than the temperature required to release them after they have been stabilized by a ligation event, All of the foregoing documents are incorporated herein by reference.
Others have recognized these limitations and tried to overcome them by providing means to accomplish multiple cycles under isothermal conditions. Examples of this are 3SR (Kwoh, D. Y. et al., Proc. Nat. Acad. Sci. USA 86; 1173-1177) and NASBA (Kievits, T. et al., 1991 J. Virol. Methods 35; 273-286, the contents of each of which is incorporated herein by reference). Each of the preceding systems has the limitation of a necessity for the introduction of an RNA promoter into the structure of the nucleic acid being amplified. Consequently, there is also a limitation that these systems are dependent upon a cycling reaction between DNA and RNA forms of the sequence of interest. A dependency upon the production of an RNA intermediate introduces a limitation of susceptibility to RNases, enzymes that are ubiquitous in the environment and are frequently present in biologically derived specimens. In addition, the nature of the design of these amplification systems has the further limitation that they require the presence of four distinct enzymatic activities: DNA polymerase, Reverse Transcriptase, RNase H and RNA polymerase. In the TMA reaction, these activities are provided by the Reverse Transcriptase and RNA polymerase enzyme whereas in 3SR and NASBA they are provided by Reverse Transcriptase, RNase H and RNA polymerase enzymes. Each of these activities is required for the system to be functional, and as such, there is a necessity for the manufacturer to test and titrate each function individually, thereby increasing the cost compared to systems that utilize a single enzyme activity. In addition, at a minimum, at least two different enzymes have to be used to provide all the necessary functions, thus rendering these systems more expensive than those that utilize a single enzyme. Furthermore, these systems require ribonucleotides as well as deoxyribonucleotides to be present as reagents for the reactions. The presence of multiple activities also creates more steps that are vulnerable to being inactivated by various inhibitors that may be present in biological specimens.
In the Strand Displacement Amplification method described by Walker et al. (Proc. Nat. Acad. Sci. U.S.A. 1992, 89; 392-396, incorporated herein by reference), isothermal amplification is carried out by the inclusion of a restriction enzyme site within primers such that digestion by a restriction enzyme allows a series of priming, extension and displacement reactions from a given template at a single temperature. However, their system has the limitation that besides the basic requirement for a polymerase and substrates, three additional elements are required in order to carry out their invention. First, there is a necessity for the presence of appropriate restriction enzyme sites at the sites where priming is to take place; secondly, there is a necessity for a second enzyme, a restriction enzyme, to be present, and lastly there is a necessity for specially modified substrates, such as thio derivatives of dNTPs to be present. A variation of this method has been described (U.S. Pat. No. 5,270,184, incorporated herein by reference) where the limitation of a necessity of a restriction enzyme site in the target has been eliminated by the use of a second set of primers that are adjacent to the primers with the restriction enzyme sites. However, in this variation, a system is described that has a new limitation of a requirement for a second set of primers while retaining the other two limitations of a need for a restriction enzyme and modified substrates.
Temperatures used for the various steps of full cycle amplification are dictated by the physical constraints that are intrinsic to each step. As such, in prior art the temperature used for complete displacement of extended strands from templates is typically around 92-95° C. This high temperature has been used to insure an adequate efficiency of separation such that an extended strand can be used as a template for subsequent reactions. When PCR was first described, the polymerase was derived from E. coli and as such, there was essentially complete thermal inactivation of the polymerase after each denaturation step that required the addition of more enzyme (Saiki et al., 1985 Science 230; 1350-1354). This problem was addressed by the use of a DNA polymerase from a thermophilic bacterium, T. aquaticus, in PCR reactions (Saiki et al., 1988 Science 239; 487-491). Each of the foregoing Saiki publications is incorporated herein by reference. Due to its inherent heat stability, enzyme was continuously present throughout the PCR cycles and no further additions were required. Since that time, polymerases from other thermophiles have also been isolated and used in full cycle reactions. However, even though they are more robust in their resistance to thermal inactivation, these polymerases all suffer from a limitation of having a certain level of inactivation after each denaturation step that is dictated by a half-life for that particular enzyme at the temperature used for denaturation. Also, the high denaturation temperature can also decrease the levels of dNTP substrates by hydrolysis and lead to inactivation of proteins that may be added to supplement the efficiency or specificity of the reaction.
Full cycle PCR conditions have been modified such that lower denaturation temperatures could be used. Auer et al. (1996, Nucl. Acids Res 24: 5021-5025, incorporated herein by reference) have described a procedure that used dITP, a natural neutral analogue of dGTP. By this substitution, they succeeded in avoiding amplification of double-stranded DNA that may be present in their samples and only amplified RNA targets. By no means is there recognition or appreciation of a utility for DNA targets. In fact, they teach away since their purpose is to avoid the use of DNA targets as templates. Their teachings have a limitation that the substitution of dITP also necessitated a compensatory decrease in the temperatures used for the annealing (50° C.). In addition, the art described by Auer et al. relies upon the use of a nucleotide analogue that is known for a lack of discrimination for base pairing, thereby introducing the possibility of random variations being introduced into the sequence being amplified. When these alterations are in the primer binding area, they may cause problems in priming efficiency and when they are in sequences between the primers, they may introduce difficulties in detection probes being able to bind efficiently. The present invention is capable of using bases that exhibit normal levels of base pairing discrimination thereby avoiding the mutagenic events that are part of the previous art.
Determination of the nucleic acid sequence of genes and genomes is a major activity in both commercial and non-profit laboratories. The two basic systems that have been employed for this purpose are the base specific cleavage method described by Maxam and Gilbert (Proc. Nat. Acad. Sci. U.S.A. 1977, 74, 560-564) and the dideoxy method described by Sanger et al. (Proc. Nat. Acad. Sci. U.S.A. 1977, 74, 5463-5467). Both of the foregoing classical papers are incorporated herein by reference. Due to its ease of use, the latter method is more commonly used. Both of these methods initially relied upon radioactive substrates for obtaining sequence information. For Maxam and Gilbert sequencing, this was most commonly carried out by end-labeling each strand and then separating each labeled end. For Sanger sequencing, either the primer is labeled or radioactive dNTPs are incorporated during strand extension. Sequence data was produced by autoradiographic determination of the position of radioactively labeled DNA bands of various lengths that had been separated by electrophoresis through a polyacrylamide gel.
In more recent years, sequencing methods have been improved by the substitution of non-radioactive labels. Non-radioactive labeling, potential positions for these labels and applications of their use are disclosed by Engelhardt et al., in U.S. Pat. No. 5,241,060, which was originally filed in 1982. Such labels can be in the oligo primer or in the substrates used for synthesis, i.e. the dNTP or ddNTP nucleotides. Signal generating moieties can act directly as exemplified by the use of fluorescently labeled primers (Beck et al., Nucleic Acids Res. 1989, 17; 5115-5123) or indirectly as exemplified by the use of biotin labeled primers (Ansorge et al., J. Biochem. Biophys. Methods 1986, 13; 315-323). In addition, biotinylated nucleotides could be incorporated during limited primer extension (Sequenase Images. TM. Protocol Book 1993 United States Biochemical Corporation, Cleveland, Ohio). The foregoing four documents are incorporated herein by reference. A limited extension is required to standardize the amount of band-shifting caused by the modification in the nucleotides.
However, primer labeling has the limitation that there can be secondary structure or problematic sequences in the template strand that can cause inappropriate chain termination events that create ambiguities in the proper base assignment for that position. Incorporation of labeled dNTPs during the extension of the primer also suffers from this limitation. This limitation is valid regardless of whether radioactive or non-radioactive labels are used.
This limitation has been circumvented by the choice of the chain terminator nucleotide itself as the source of the label. This has been described by Hobbs and Cocuzza in U.S. Pat. No. 5,047,519 and by Middendorf et al., in U.S. Pat. No. 4,729,947 for fluorescently labeled ddNTPs and by Middendorf et al., in U.S. Pat. No. 4,729,947 for biotin labeled ddNTPs that were later marked by fluorescent avidin. (For further reference, refer to U.S. Pat. Nos. 5,027,880; 5,346,603; 5,230,781; 5,360,523; and 5,171,534.) Each of the foregoing seven patents is incorporated by reference into this application. By this method, signals will be generated by strands that have incorporated a chain terminator. The presence of strands that have been terminated without the incorporation of a terminator nucleotide is now irrelevant since they are incapable of signal generation. However, this method has the limitation that the presence of additional chemical groups that provide signal generation produce steric or other inhibitory problems for the polymerase directed incorporation of the labeled terminator nucleotide, thereby decreasing the efficiency of the reaction (Prober et al. in U.S. Pat. No. 5,332,666, incorporated herein). It has also been suggested that biotinylated dideoxynucleotides could be used to provide signal generation, but these modified terminators were predicted to share the same limitations as their fluorescenated counterparts, i.e. difficulty in incorporation by most commonly used polymerases (S. Beck 1990 Methods in Enzymology 184; 611, also incorporated herein). Some compensation for this inefficiency of incorporation can be achieved by increasing the amounts of polymerase in the reaction and/or by increasing the amount of template DNA being copied. These compensatory steps suffer the limitation of increased costs associated with higher amounts of an expensive enzyme, DNA polymerase, or with preparation of adequate amounts of high quality template.